ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent/Changes
(Generate patch)

Comparing AnyEvent/Changes (file contents):
Revision 1.175 by root, Sat Jul 19 22:10:47 2008 UTC vs.
Revision 1.571 by root, Tue Jul 30 23:14:32 2013 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.22 3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
10
11TODO: expose default formatting implementation (ft &c) in AE::log, also allow overridign fatal exit
12TODO: José Micó (3.0K) 1 inbox Error connecting to hosts resolved from /etc/hosts
13TODO: Daniel Carosone (5.2K) 1 inbox AnyEvent::IO vs AnyEvent::AIO (doc patch)
14TODO: maybe implement env variable to give hosts precedence
15 - uts46data.pl couldn't be found due to wrong naming of the file
16 (reported by Fulko Hew).
17 - AnyEvent::Debug shell can now run coro shell commands, if available.
18 - t/63* tests were wrongly in MANIFEST.
19 - mention IO::AIO as optional dependency.
20 - kernel.org's finger server went MIA, switch to freebsd.org and
21 icculus.org.
22 - handle lone \015's properly in AE::Handle's default line read
23 (reported by various people).
24 - hard-disable $^W in most tests, it generates too much garbage output.
25 - use a (hopefully) more future-proof method to emulate common::sense.
26
277.04 Thu Nov 15 02:23:58 CET 2012
28 - debugging code left in.
29
307.03 Thu Nov 15 02:16:11 CET 2012
31 - AnyEvent::Socket::inet_aton did not work when dns resolution was
32 used to find the addresses (analyzed and patch by Maxime Soulé).
33 - fix a memory leak in the /etc/hosts lookup code when hosts don't
34 resolve and are not in hosts (analyzed by sten).
35
367.02 Tue Aug 14 04:27:58 CEST 2012
37 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
38 program by Yorhel).
39 - verified that AnyEvent::Socket follows RFC5952.
40 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
41
427.01 Sun May 13 01:03:17 CEST 2012
43 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
44 not available, instead of throwing an exception.
45 - use File::Spec to get the tmpdir in t/*, to avoid needless
46 failures on (most, not mine :) windows boxes.
47 - new handle read types: tls_detect and tls_autostart.
48
497.0 Fri Apr 13 06:33:30 CEST 2012
50 - child watchers are broken in POE 1.352 (also many earlier
51 versions) and there seems to be no way to work around it, as POE
52 itself is inherently racy. Document this regression and add a delay
53 in t/68_poe_03_child.t for the time being.
54 - new module AnyEvent::IO, that is a frontend to either a pure-perl
55 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
56 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
57 - load /etc/hosts only when DNS has no answer.
58 - stat /etc/hosts on every access and reload it if it changed.
59 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
60 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
61 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
62 a side effect, they now use true appending as opposed to libc
63 appending, and the file might not have been opened when the function
64 returns.
65 - the default logging level was not properly documented in a variety of
66 places, this has hopefully been rectified.
67 - updated uts46data.pl for unicode 6.1.0.
68 - made log messages generated by AnyEvent submodules not
69 include the package name anymore, as it will be logged
70 by default already.
71 - upgrade to the trick used by common-sense 3.5 to work
72 around extra warning torture/breakage under perl 5.15.x.
73 - log messages by anyevent are now ucfirst, are usually full
74 sentences and do no longer include the package name.
75 - the storable read type would sometimes throw an exception instead
76 of causing EBADMSG (analyzed by Maxime Soulé).
77 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
78 keysize (reported by Paul Howarth).
79 - documented Rocco fabricating statements about AnyEvent and me.
80
816.14 Tue Jan 31 20:00:24 CET 2012
82 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
83 probably an editing mistake (reported by Darin McBride).
84
856.13 Thu Jan 12 07:27:01 CET 2012
86 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
87 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
88 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
89 regex-matching inside the callback would change the parameter, despite
90 $1 being dynamically scoped per-block. this perl bug is now being
91 worked around (testcase by Cindy Wang).
92
936.12 Mon Dec 12 13:21:10 CET 2011
94 - $! was clobbered by subsequent calls in tcp_connect, due to
95 the postpone not saving/restoring it (analyzed by Richard Garnish).
96
976.11 Tue Nov 22 10:36:05 CET 2011
98 - Tk cannot create windows when tainted unless you set a title, so
99 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
100 - escape any nonprintable/nonascii characters when stringifying
101 backtraces.
102 - log the reason loading an interface module fails to load at level debug.
103
1046.1 Tue Oct 4 19:44:30 CEST 2011
105 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
106 and above), and some messages inside AnyEvent have been elevated
107 to higher log levels to print by default.
108 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
109 - work around yet another signal race bug, newly introduced by POE. My
110 how that module sucks.
111 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
112 support ipv6-only operation even with defaults (also use ::1 instead
113 of 127.0.0.1).
114 - give better guidelines on how to chose a logging level.
115 - AnyEvent::Log can now cap (limit) log message priority by context.
116 - AE::log will now save and restore $! and $@, so callers don't have to.
117 - AE::Strict now checks that registered fds are still valid at regular
118 intervals.
119 - skip frequently called 'require' calls if the module appears loaded.
120 - increase signal latency test timer considerably for poor little osx,
121 which can't be bothered with basic posix functionality such as signal
122 delivery.
123 - mention some debugging aids in the tutorial.
124 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
125 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
126 of the context as title.
127 - AnyEvent::Strict would not restore read-write access to $_ when a callback
128 died - this is a user bug, but since AE::Strict exists to help dbeugging...
129 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
130 to work with it - the author is well aware of the issues but refuses
131 to fix them. This extreme step has been taken because the module actively
132 attacks the goals of AnyEvent and makes it harder for module authors to use
133 AnyEvent.
134
1356.02 Fri Aug 26 20:08:31 CEST 2011
136 - log_to_syslog now accepts facility strings.
137 - log_to_syslog would not allow facilities - "$facility|$level"
138 *literally* means that string. sheesh.
139 - the CPAN parser got confused and thought we provide Sys::Syslog, try
140 to work around.
141 - renamed AnyEvent::Impl::FLTK2 to FLTK.
142
1436.01 Fri Aug 26 07:04:11 CEST 2011
144 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
145 the TTL as fourth element - this affects ->resolve, ->request
146 and ->dns_unpack indirectly. Convenience functions such as
147 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
148 AnyEvent::Socket.
149 - INCOMPATIBLE CHANGE: backend authors now should not implement
150 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
151 methods.
152 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
153 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
154 - DEPRECIATION: please change
155 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
156 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
157 in your programs, for future compatibility.
158 - *blush* idle watcher emulation was borked.
159 - AnyEvent::Strict now checks the AE::xxx functions as well.
160 - querying too long domain names would barf deep inside AnyEvent::DNS - since
161 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
162 for too long domain names explicitly and passes an undef to the callback
163 instead (reported by James Bromberger).
164 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
165 - new functions: AnyEvent::postpone, AnyEvent::log.
166 - new module: AnyEvent::Log, for simple logging needs.
167 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
168 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
169 a debug shell.
170 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
171 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
172 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
173 /etc/hosts can be found.
174 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
175 configuration.
176 - env variables can now be specified using AE_xyz instead of
177 PERL_ANYEVENT_xyz, subject to some rules.
178 - lots of new AnyEvent::Debug functionality.
179 - AnyEvent::Strict now checks for modifications of $_ by the callback
180 (by setting it to readonly while executing the callback, if possible).
181 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
182 Invest extra overhead to work around this peculiar behaviour.
183 - split the pure perl backend into a loop module and an impl
184 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
185 - some read types (regex, netstring, json, storable) could
186 errornously keep an AE::Handle object alive on their own
187 (reported by Mohammad Toossi).
188 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
189 supports /etc/hosts overrides.
190 - add more workarounds around child watcher bugs in IO::Async
191 and (newly introduced) in POE.
192 - use glib child watchers and try to work around its limitations,
193 if possible (based on a patch by Kevin Ryde).
194 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
195 - many read types in AnyEvent::Handle would malfunction during
196 transitions to TLS because of a "delete $self->{rbuf}".
197 (reported by Gerald Galster).
198 - improve condvar blocking wait performance (probably only noticable
199 with faster backends), also make it easier for event loops that
200 do not support blocking waits.
201 - PERL_ANYEVENT_MODEL now also allows full module names, although
202 the usefulness of this feature is not yet clear.
203 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
204 - add t/80_ssltext.t.
205 - no longer use AUTOLOAD for the initial loading (for no very good reason).
206 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
207 - do not run event loop detection multiple times when programs call their cached
208 copies of AnyEvent::detect.
209 - parse_hostport now accepts absolute paths als unix domain sockets.
210 - suppress the idiotic warning inside POE by patching the POE::Kernel
211 run flags instead of asking the user to jump through contortions.
212 - add Task::Weaken as dependency (as recommends only at this time).
213 - replace PF_UNSPEC by 0, for good old minix.
214
2156.0 (not officially released, but was on CPAN for about 15 minutes :/)
216
2175.34 Thu May 12 09:59:41 CEST 2011
218 - AE::TLS didn't support both cert_file and key at the same time,
219 due to some copy&paste bug (reported and analyzed by vti).
220 - improve AE::Handle timeout documentation slightly.
221
2225.33 Mon Feb 7 21:51:41 CET 2011
223 - never appeared on CPAN, for unknown reasons.
224 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
225 object when it's already gone (analyzed by Ryan Bullock).
226 - update idna algorithm(s) to UTS#46 version 6.0.0.
227 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
228 uts46 says it's an error, and testsuite makes more sense.
229 - remove leading dots in idn_to_ascii - testsuite says to remove,
230 uts46 says it's an error, and testsuite makes more sense.
231 - fixed a bug in idn ignored character processing.
232 - passes the uts46 6.0.0 testsuite except for three cases which
233 apparently are bugs in the testsuite itself.
234 - passes the uts46 6.0.1 testsuite except for eleven cases which
235 apparently are bugs in the testsuite itself.
236
2375.31 Sun Jan 23 11:49:19 CET 2011
238 - work around a POE bug where POE would first poll for new events and
239 invoke handlers BEFORE handling the message queue.
240 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
241 - stop_read now does just that in TLS mode. see its
242 documentation for caveats.
243 - update to common::sense 3.4 warning set (some warnings in perl impose
244 an arbitrary 32kb limit on regex matches, which affects
245 AnyEvent::Handle).
246 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
247 - new experimental AE::Handle::resettls method.
248
2495.3 Fri Dec 31 05:48:13 CET 2010
250 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
251 issues with IO::Async have been solved in that version, including
252 the availabilty of a default loop!
253 - implement new wbuf_max parameter for AnyEvent::Handle.
254 - added Cocoa::EventLoop backend by Daisuke Murase.
255 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
256 - clarify handle on_prepare documentation.
257 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
258 to avoid leaking them to any child processes.
259 - try to force files into the arch-dependend lib directory, for the
260 half-"turly obsessive compulsive person"(s) that depend on it.
261
2625.29 Sun Dec 5 10:49:21 CET 2010
263 - convert EV backend to EV 4.00 API (so better upgrade EV too).
264 - AE::Handle: implement dynamic read_size adjustment, add
265 max_read_size parameter, reduce default read_size to 2048.
266 - add some FAQ entries for servers and/or on_eof vs. on_error.
267 - work around OS bugs (cygwin again) when lingering in AE::Handle.
268 - reduce memory usage slightly when connect was used in AE::Handle.
269 - wrok around more segfaults in Tk.
270 - document tls_ctx => undef as valid.
271 - detect WNOHANG value at compiletime.
272 - include some probably linux-only support for building AnyEvent
273 as part of the perl core.
274 - improve compatibility of t/08_idna.t to perl 5.8.
275 - make AnyEvent::Strict truly optional (in case somebody builds a
276 minimal perl...).
277
2785.28 Wed Oct 13 04:14:23 CEST 2010
279 - due to a glitch, AnyEvent's internal getprotobyname wasn't
280 actually used in tcp_server.
281 - implement AnyEvent::CondVar->new method, which was documented but
282 missing.
283 - support autoloading of read types in unshift_read.
284 - AnyEvent::Strict: do the test for invalid fh arguments differently,
285 resulting in more informative error messages.
286 - new AnyEvent::FAQ manpage.
287 - clarify that you can actually call push_read etc. in on_read.
288 - put keys and certs in /tmp, not into the current directory,
289 and document this in the SECURITY CONSIDERATIONS sections
290 of AnyEvent::TLS.
291 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
292 requested by Ron Isaacson.
293 - document $AnyEvent::DNS::RESOLVER variable.
294 - incorporated major typo patches by Abhijit Menon-Sen.
295 - AnyEvent::Handle now croaks when negative timeouts are passed.
296 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
297 undocumented, but who knows when they might come in handy.
298
2995.271 Tue Jun 8 12:05:46 CEST 2010
300 - backport to perl 5.8.x.
301
3025.27 Sun Jun 6 12:12:05 CEST 2010
303 - postpone differently in AnyEvent::Socket now, as
304 when not, canceling the connection attempt might fail
305 (found by Felix Antonius Wilhelm Ostmann).
306 - explicitly check for non-stream sockets in AE::Handle, too many
307 clueless people fell into the trap of this somehow working.
308 - simplified and reworked the "OTHER MODULES" section.
309 - better/more condvar examples.
310
3115.261 Wed Apr 28 16:13:36 CEST 2010
312 - AF_INET6 was not properly used from Socket6 during configuration
313 time (found by Andy Grundman).
314
3155.26 Mon Apr 12 04:49:35 CEST 2010
316 - don't generate spurious readyness notifications when select
317 returns EINTR in the pure perl backend (analysed and testcase
318 by Bas Denissen).
319 - give same examples in SYNOPSIS sections of AnyEvent and AE.
320 - provider faster implementations for AE::time/now/now_update for
321 EV and Perl backends.
322 - provide faster AE::cv and AE::time implementations for all
323 backends.
324 - no longer support lower/mixed-case signal names (this was
325 never documented nor universally supported).
326 - some more memory µ-optimisations, and somewhat less messy
327 code to cope with both AnyEvent and AE APIs.
328
3295.251 Sat Mar 13 00:58:21 CET 2010
330 - make test failed if it was run before make install on perl <=
331 5.8.8, and older versions of perl make our live very difficult,
332 so write constants.pl during Makefile.PL time.
333
3345.25 Sat Mar 13 00:23:14 CET 2010
335 - fix a race condition in AnyEvent::Handle that would cause
336 a "bio_read: unsupported method" error (in _another_
337 ssl connection) after user code threw an exception.
338 - added AnyEvent::Handle->destroyed method.
339 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
340 - the AnyEvent::Util::fh_nonblocking function worked only
341 by lucky accident on win32.
342 - smaller and faster AnyEvent::Util::fh_nonblocking.
343 - when the (required!) Time::HiRes module is missing, AnyEvent
344 did not fall back to built-in time properly.
345 - do not load Fcntl at runtime, saving memory and loading time.
346 - precompile a number of constants and use them instead of runtime
347 detection and eval.
348 - free detection code after detection and similar memory optimisations.
349 - Perl backend timer interval best effort drift has been improved
350 (same algorithm as EV).
351 - update unicode idna mapping table.
352
3535.24 Tue Jan 5 11:39:43 CET 2010
354 - cygwin never reports errors from failed connects. 1.5 just gives
355 you continous readyness and EAGAIN, 1.7 is even more broken and
356 just hangs. work around both issues in a cygwin-specific hack.
357 - improve idle watcher documentation slightly.
358
3595.23 Sun Dec 20 23:48:00 CET 2009
360 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
361 - implement punycode_encode/decode, idn_nameprep,
362 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
363 - provide $AE::VERSION.
364 - removed traces of "no strict 'refs'".
365
3665.22 Sat Dec 5 03:51:13 CET 2009
367 - downgrade-or-fail in AnyEvent::Handle::push_write, to
368 diagnose encoding failures earlier and more succinctly.
369 (this works around bugs in perl, throwing away encoding info
370 when passing scalar data to extensions).
371 - add more examples to AnyEvent::Socket manpage.
372 - upgrade internal warning set to the same as common::sense 2.03.
373 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
374 (John Beppu).
375 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
376 - add google public dns servers to fallback server set and make sure
377 we load-balance properly between the three sets. also add all
378 fallback dns servers, not just a random one, to each dns config.
379
3805.21 Thu Nov 19 02:48:47 CET 2009
381 - fix a problem where socket constants were called with parameters
382 (spotted by David Friedland).
383 - fork_call never use'd POSIX (reported by Daisuke Maki).
384 - improve perl 5.6 compatibility further (but it still won't work
385 unless you rip out everything but the core).
386 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
387 uses 10MB of RAM, but doesn't open a console window. *sigh*.
388
3895.202 Wed Oct 14 22:35:44 CEST 2009
390 - AnyEvent::DNS would unexpectedly clobber $_ under windows
391 (analysed by Matthias Waldorf).
392 - AnyEvent::Handle::run_cmd can now pass the PID of the
393 newly-created process, which is much less useful than it might
394 sound (based on patch by Yann Kerherve).
395
3965.201 Tue Sep 29 12:09:25 CEST 2009
397 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
398 (reported by Torsten Foertsch).
399 - common::sense 2.0 could cause tcp_server to throw an exception
400 (analysed by elmex).
401
4025.2 Mon Sep 14 07:04:49 CEST 2009
403 - INCOMPATIBLE CHANGE: do no longer support register_read_type
404 and register_write_type in AnyEvent::Handle, instead support
405 package names (the facility was mostly abused).
406 - implement "packagename-as-read/write type" support in
407 AnyEvent::Handle.
408 - AnyEvent::Handle: new options "keepalive" and "oobinline".
409 - oobinline set by default to avoid security issues.
410 - the pure-perl event loop backend wrongly detected times() fallback
411 support (spotted by Pavel Boldin).
412
4135.12 Tue Sep 1 20:26:50 CEST 2009
414 - be more lenient when parsing resolv.conf files, as some people
415 use hashmarks as comment indicator inside directives (reported
416 by Michael S. Fischer).
417 - use same set of warnings as common::sense 2.0.
418 - fix a potential 32 bit overflow issue due to perl having problems
419 with large hex constants in the Perl backend.
420 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
421 by default.
422 - fix some prototypes.
423 - (optionally) require Async::Interrupt 1.02.
424 - replace opendns fallbacks by verizon/level3 public dns servers.
425
4265.112 Fri Aug 21 13:59:04 CEST 2009
427 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
428 using starttls with an illegal/uncreatable tls context.
429 - port to broken win32 perls.
430
4315.111 Sun Aug 16 18:44:36 CEST 2009
432 - perl errornously requires sockaddr_un structures returned by the
433 kernel to have a specific length. work around this bug by padding
434 sockaddr_un structures (found to fail on OS X, as perl only uses
435 the correct code for GNU/Linux, but likely to fail on other
436 OSes as well).
437
4385.11 Wed Aug 12 17:49:37 CEST 2009
439 - fixed a bug in run_cmd with input from filename.
440 - create a json coder object when none is given in write direction
441 to reduce dependency on JSON.
442
4435.1 Tue Aug 11 03:17:32 CEST 2009
444 - tcp_connect now reports the correct error in $!, not an unrelated
445 one (was broken since 4.91).
446 - AE::Handle did not properly clear rtimeout/wtimeout watchers
447 when retrying a tcp_connect.
448 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
449 - fixed wrong documentation in AE::Socket.
450
4515.01 Mon Aug 10 03:16:32 CEST 2009
452 - last release broke EV child watchers :/.
453
4545.0 Sun Aug 9 17:32:09 CEST 2009
455 - officially introduce and document the AE API.
456 - lowering the timeout at runtime did not have immediate effect
457 in AnyEvent::Handle.
458 - AE::Handle now has separate and independent read and write timeout
459 settings.
460 - probe for Guard module when AnyEvent::Util is loaded, not at
461 runtime (Event doesn't like eval "use" from callbacks, and it avoids
462 exporting the probe function).
463 - Impl::IOAsync's io watchers did not use the filehandle emulation
464 code to work around it's watcher limits.
465 - work around IO::Async's problems with overloaded objects.
466 - try to disable wearnings inside IO::Async, it's just too buggy.
467 - do no longer use Event by default, only EV or Perl, Event is too buggy.
468 - hackishly provide fast AE:: interface when EV or Perl is the backend.
469 - inherit VERSION from AnyEvent::VERSION in most modules.
470
4714.91 Thu Aug 6 15:42:45 CEST 2009
472 - AE::Handle::starttls could get out of sync when the read buffer
473 already contains some TLS handshake.
474 - AE::Handle did not properly free the TLS session, causing bigger
475 memory leaks in Net::SSLeay than expected (reported by toaster).
476 - AE::Socket::tcp_connect will now artificially delay invoking
477 the callback to avoid returning after invoking the callback.
478 - convert many internal watcher uses to AE API.
479
4804.9 Sat Aug 1 11:07:01 CEST 2009
481 - Glib silently fails when registering a timeout with
482 a negative value, so avoid that.
483 - call condvar callback immediately when it is set
484 after the condition is already signalled.
485 - check rbuf_max condition only after trying to consume
486 data.
487 - normalise signal numbers to names when using the ae's
488 signal handling, but do not document this (yet).
489 - pure perl signal emulation did not properly set nonblocking
490 mode on the signal pipe on win32 (this is unlikely to have caused
491 any issues).
492 - new module: AnyEvent::Debug.
493 - AnyEvent::Strict now checks that a signal was specified
494 by name (not name or number).
495 - reduce memory footprint in typical cases by ~50kb by
496 only compiling signal code when necessary.
497 - add AnyEvent::Handle->rbuf_max.
498 - grab the AE namespace for future new API, implement stubs for
499 said future/alternative API.
500 - new function: AnyEvent::Socket::format_hostport.
501
5024.881 Tue Jul 28 12:51:53 CEST 2009
503 - work around a bug in local in pre-5.10 perls, causing
504 AnyEvent::Handle to recurse when it should not
505 (analysed by elmex).
506
5074.88 Tue Jul 28 04:04:37 CEST 2009
508 - re-bless the handle into a dummy package after calling
509 AnyEvent::Handle::destroy, so the user does not need to check
510 for errors after every push_write etc.
511 - do not attempt to run t/02_signals.t on obviously broken
512 platforms.
513
5144.87 Sun Jul 26 02:06:16 CEST 2009
515 - do not attempt to linger when there is no longer a valid fh.
516 - enforce tls mode to be either accept or connect, do not simply
517 segfault in Net::SSLeay.
518 - AnyEvent::Handle can now call tcp_connect itself (new parameters
519 connect, on_prepare, on_connect and on_connect_error). Updated
520 tutorial accordingly.
521 - add AnyEvent::Impl::Irssi backend.
522
5234.86 Mon Jul 20 23:52:29 CEST 2009
524 - since the verbose warning is not enough, explicitly document
525 that versions before 1.33 of Net::SSLeay are not secure.
526 - work around signal handling races in Event and (...) Event::Lib.
527 - try to align signal-race timer to full-second boundaries.
528 - work around Tk not liking negative timeouts.
529 - don't complain of different grades of environmental unfriendlyness
530 in IO::Async.
531
5324.85 Sat Jul 18 06:16:14 CEST 2009
533 - nail the signal race problem in perl once and for all
534 (see $AnyEvent::MAX_SIGNAL_LATENCY).
535 - take advantage of Async::Interrupt if it is available.
536 - load Time::HiRes and Guard modules on demand only.
537 - add optional/recommended modules section to AnyEvent
538 documentation.
539 - reduce memory usage considerably (and reduce startup penalty)
540 by not using "strict", "warnings" and "overload" modules.
541 - work around buggy windows/openbsd perls and provide EBADMSG
542 and EPROTO ourselves when missing.
543 - improve perl 5.6 compatibility of the core event loop.
544 - made Net::SSLeay version 1.33 a soft requirement.
545
5464.83 Fri Jul 17 16:56:26 CEST 2009
547 - implement AnyEvent::Socket::getprotobyname.
548 - AnyEvent::CondVar's will now detect recursive blocking
549 waits and will croak, as too many people fall into
550 this trap.
551 - AnyEvent::Handle will now call ->destroy on itself after
552 executing the on_error callback, instead of doing some
553 half-baked internal shutdown, for fatal errors.
554 - clarify on_eof behaviour w.r.t. the read queue and
555 on_read callbacks.
556 - ignore some possible spurious wake-ups in tcp_connect.
557
5584.82 Sat Jul 11 00:34:55 CEST 2009
559 - POE and Event backends didn't accept some callable objects as
560 callbacks.
561 - use Config module instead of POSIX module to detect signal names
562 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
563 module doesn't even have all POSIX signals :/.
564 - use more workarounds around the many refcnt/corruption bugs in
565 Event::Lib.
566 - work around a race condition in perl's select, causing t/03_child.t
567 to rarely fail.
568
5694.81 Thu Jul 9 10:30:30 CEST 2009
570 - AnyEvent::Handle didn't properly diagnose write errors
571 (it expected -1 from syswrite, how lame... :).
572 - support file descriptors in addition to file handles
573 in AnyEvent->io.
574 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
575 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
576 and PERL_ANYEVENT_CA_PATH.
577 - provide a sensible synopsis section for AnyEvent::TLS.
578 - add a "supported backends" section to the manpage.
579 - added simple io watcher test to testsuite, using a
580 portable_socketpair.
581 - tried to improve the stability of the Event::Lib backend,
582 YMMV.
583
5844.8 Mon Jul 6 23:45:16 CEST 2009
585 - AnyEvent::DNS did not properly follow CNAME records with
586 uppercase targets.
587 - AnyEvent::DNS would errornously return AAAA records
588 with v4 mapped addresses (a faulty record) as ipv4 addresses,
589 causing AnyEvent::Socket to throw an exception.
590 - added new module AnyEvent::TLS for easier SSL/TLS context
591 creation, with many options including hostname verification,
592 secure default configuration, lots of documentation and,
593 predefined diffie-hellman keys for perfect forward security
594 and much more. get it while it's still fresh!
595 - use AnyEvent::TLS in AnyEvent::Handle for context management.
596 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
597 so AnyEvent::Socket users have smaller memory footprint
598 in the common case.
599 - add AnyEvent::Handle->push_shutdown method.
600 - add an additional error message parameter to AnyEvent::Handle's
601 on_error callback (for TLS, $! is still available).
602 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
603 - make AnyEvent::Handle more robust against early conenction
604 failures (during new), and return C<undef> in that case
605 from the constructor.
606 - AnyEvent::Handle will now try to load only JSON::XS first,
607 then fall back to JSON.
608 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
609 and because it was documented that way.
610
6114.451 Fri Jul 3 00:28:58 CEST 2009
612 - do not clear rbuf when shutting down an AnyEvent::Handle
613 object - doing so breaks AnyEvent::HTTP.
614
6154.45 Mon Jun 29 22:59:26 CEST 2009
616 - a write error could cause AnyEvent::Handle to create
617 an I/O watcher with an undefined $fh.
618 - special-case mapped ipv4 addresses in both
619 AnyEvent::Socket::format_address and parse_address,
620 to treat them just like ipv4 addresses.
621 - updated and overhauled the AnyEvent::Intro doc.
622 - implement AnyEvent::Socket::format_ipv4/ipv6.
623 - slightly speed up portable_pipe/socketpair.
624 - expand condvar begin/end documentation.
625
6264.42 Fri Jun 26 08:32:18 CEST 2009
627 - preliminary and neccesarily incomplete support for IO::Async.
628 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
629 case it was set to IGNORE, to ensure we can catch child statuses
630 even when the calling env acts stupidly.
631 - updated benchmarks with IO::Async, which performs very well.
632
6334.412 Wed Jun 24 01:35:57 CEST 2009
634 - support an "untaint" attribute for AnyEvent::DNS
635 and set it on the default resolver.
636 - implement and document AnyEvent::Impl::Perl::loop.
637 - remove all anyevent-env variables from %ENV
638 when running in tainted mode.
639 - mention and extend the IO::Lambda benchmark.
640
6414.411 Sun Jun 7 18:48:02 CEST 2009
642 - do not try to use F_SETFD on windows.
643
6444.41 Thu May 14 06:40:11 CEST 2009
645 - work around issues in older perls (5.8.0?) when
646 a signal handler is deleted from the %SIG hash.
647 - use POSIX::_exit in child test, to avoid running
648 destructors.
649 - speed up CHLD handling by relying on SIGCHLD being
650 synchronously delivered, even when we roll our own
651 implementation.
652 - AnyEvent::DNS: add the "dname" resource record name
653 for cosmetic reasons.
654
6554.4 Sun Apr 26 20:12:33 CEST 2009
656 - implemented idle watchers, where applicable.
657 - AnyEvent->time died when Event backend was in use.
658 - fix a memleak in the Tk backend.
659 - sped up Tk timer handling.
660 - clip negative "after" values to 0 in AnyEvent::Impl::Event
661 to avoid spamming out warning messages.
662 - fix Qt timers without interval.
663 - avoid Qt zero-timeout special case, in old libqt's.
664
6654.352 Mon Apr 20 16:31:11 CEST 2009
666 - fix AnyEvent::Strict error messages for child watchers.
667 - fix/update Makefile.PL "recommends" versions.
668 - add AnyEvent->now_update.
669
6704.351 Sat Apr 11 07:56:14 CEST 2009
671 - actually make the signal pipe work under win32.
672 - localise $! in signal handler to avoid changing $!, although
673 perl itself does not seem to save/restore errno either.
674 - set the cloexec flag on the signal pipe (normally set by perl too).
675
6764.35 Fri Mar 27 11:48:20 CET 2009
677 - event models relying on AnyEvent's signal watcher emulation did
678 invoke the callback asynchronously, contradicting documentation
679 and causing signals to get lost (this includes AnyEvent's own
680 event loop). AnyEvent now uses the standard pipe trick to make
681 callback execution synchronous to the event loop.
682 - AnyEvent::Handle didn't free TLS context data on DESTROY
683 (patch by Pavel Shaydo).
684 - work around the ever-incompatibly-changing API of MakeMaker.
685 - document that changing global variables without restoring them
686 is a bad idea in Perl (noted by Adam Rosenstein).
687 - AnyEvent::Strict now barfs if ->io is passed a file.
688
6894.34 Thu Feb 12 18:32:45 CET 2009
690 - separately buffer TLS read data, as otherwise the read queue could
691 deadlock as receiving data is not expected while draining the
692 read queue (which cna only happen with TLS).
693 - raise EBADMSG error on JSON decoding errors.
694 - fix some minor manpage bugs (reported by Maximilian Gaß).
695 - speed up select bitmask parsing quite a bit in the pure perl
696 backend.
697 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
698
6994.331 Tue Jan 6 21:07:25 CET 2009
700 - socketpair fails on many vista machines because vista has
701 completely broken accept/getpeername and getsockname functions,
702 so we provide our own socketpair emulation that kind of works
703 (AnyEvent::Util::portable_pipe).
704 - new function: AnyEvent::Util::portable_socketpair.
705 - take advantage of the Guard module if it exists.
706
7074.33 Fri Nov 21 02:35:40 CET 2008
708 - AnyEvent::Strict did errornously flag a fileno of 0 as
709 illegal.
710 - reduce memory usage and slightly speed up the pure perl backend
711 by only storing the file descriptor, not the file handle.
712 - add missing autocork method to AnyEvent::Handle (reported
713 by Adam Rosenstein).
714 - AnyEvent::DNS->resolve errornously documented an $rcode result
715 argument, but there isn't (spotted by Henrik Krohns).
716 - the naked truth about Tk - it's basically unmaintained.
717
7184.32 Mon Nov 3 22:46:32 CET 2008
719 - fix AnyEvent::Socket::resolve_sockaddr to properly support
720 unix sockets again and choose a proper default for
721 the socket type.
722 - fix call to resolve_sockaddr in tcp_connect.
723
7244.31 Thu Oct 30 04:41:48 CET 2008
725 - implemented AnyEvent::Handle->destroy method.
726 - hint about unexpected effects in TLS mode.
727 - speed up AnyEvent::Impl::Perl by using more arrays
728 instead of hashes without the slightest loss of readability :)
729 - work around a perl argument refcounting bug.
730 - with some perl patches applied, AnyEvent now seems leak-free.
731
7324.3 Fri Oct 3 09:18:43 CEST 2008
733 - AnyEvent will now install a no-op signal handler for SIGPIPE,
734 unless one has been installed already.
735 - warn about not loaidng AnyEvent::Impl::POE early enough
736 (patch by Adam Rosenstein).
737 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
738 - fix a bug in where in SSL connect mode, the
739 client would first wait for some data by the server and
740 otherwise hang. (reported and analysed in an absolutely
741 exemplary manner by Adam Rosenstein).
742 - fix a bug in where SSL EOF would not be treated
743 as stream EOF, putting the connection into a hung state.
744 (reported and analysed in an absolutely exemplary manner by Adam
745 Rosenstein).
746 - fix a potential segfault when the TLS context
747 would go missing in a read callback (e.g. due to stoptls) -
748 Net::SSLeay of course makes no type checking whatsoever.
749 - AnyEvent::Handle will not stop the read watcher in TLS mode.
750 - AnyEvent::Handle->stoptls will send a TLS close notify
751 instead of simply destroying the stream now.
752 - fix a bug where large blocks of data written in TLS mode
753 would not be sent unless triggered by receives.
754 - on_drain will now take the tls write buffer into account.
755 - SSL operations have been streamlined a bit and should be faster
756 now, more to come, though.
757 - added a FAQ section, will document nontrivial issues in there.
758
7594.234 Mon Sep 29 04:08:13 CEST 2008
760 - fix child watcher documentation: only child exits will be reported,
761 no trace events.
762 - mention SIGPIPE in AnyEvent::Handle.
763 - perl backend will now use POSIX::times as monotonic clock
764 when available and the monotonic clock option is not.
765 - run even without Time::HiRes in some cases now, but do not
766 provide sub-second accuracy (all *supported* perl versions have
767 Time::HiRes).
768
7694.233 Fri Aug 22 01:48:05 CEST 2008
770 - fix a bug in the testsuite, causing a hang.
771 - clarified lots of AnyEvent::Handle constructor arguments.
772
7734.232 Thu Aug 21 20:44:25 CEST 2008
774 - fix a bug in AnyEvent::Handle that could cause two fatal error
775 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
776 (testcase by Leon Brocard).
777 - clarified on_eof/on_error documentation a bit.
778
7794.231 Tue Jul 29 13:12:15 CEST 2008
780 - remove some debugging code left in AnyEvent::Util::fork_call
781 (and no, it's impossible to implement with the broken windows
782 perls without resource leaks or worse).
783
7844.23 Tue Jul 29 12:19:59 CEST 2008
785 - document the first parameter passed to condvar callbacks to be
786 the callback.
787 - add AnyEvent::Socket::{ntoa,aton} aliases.
788 - optimize the AE::Handle->push_read (line) for the default
789 eol marker.
790 - optimize push_read (packstring|storable) for small packets.
791 - invoke on_error callback when no on_eof callback is set.
792 - fix a bug in push_read (storable) of unknown impact.
793
7944.22 Sun Jul 20 16:34:13 CEST 2008
4 - new function AnyEvent::Socket::parse_hostport. 795 - new function AnyEvent::Socket::parse_hostport.
5 - as the bulkheads at microsoft can't even get getprotobyname reliably 796 - as the bulkheads at microsoft can't even get getprotobyname reliably
6 working on their shitty fucking broken os we need to hardcode 797 working on their shitty fucking broken os we need to hardcode
7 some common protocol numbers in AnyEvent::Socket. How can 798 some common protocol numbers in AnyEvent::Socket. How can
8 people even bother with such a pile of shit as windows. 799 people even bother with such a pile of shit as windows.
… …
187 - support service names in tcp_connect and tcp_server. 978 - support service names in tcp_connect and tcp_server.
188 - work around netbsd/openbsd bug causing unix sockets not to 979 - work around netbsd/openbsd bug causing unix sockets not to
189 become writable until empty. 980 become writable until empty.
190 - only make a dummy read for error detection when getpeername 981 - only make a dummy read for error detection when getpeername
191 returns ENOTCONN. 982 returns ENOTCONN.
192 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 983 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
193 override. 984 override.
194 - new AnyEvent::Guard::cancel. 985 - new AnyEvent::Guard::cancel.
195 986
1963.5 Sat May 17 23:17:14 CEST 2008 9873.5 Sat May 17 23:17:14 CEST 2008
197 - gracefully deal with systems not defining CLOCK_MONOTONIC 988 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
211 1002
2123.41 Sun May 11 19:53:13 CEST 2008 10033.41 Sun May 11 19:53:13 CEST 2008
213 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 1004 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
214 1005
2153.4 Sun May 11 00:29:25 CEST 2008 10063.4 Sun May 11 00:29:25 CEST 2008
216 - complete re-implementation of condvars with a hopefully much mroe useful 1007 - complete re-implementation of condvars with a hopefully much more useful
217 API (backward compatible functions exist but are undocumented). 1008 API (backward compatible functions exist but are undocumented).
218 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 1009 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
219 (still subject to change). 1010 (still subject to change).
220 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 1011 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
221 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 1012 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines